Professional Documents
Culture Documents
Markus K. Brunnermeier
LECTURE 06:
MEAN-VARIANCE ANALYSIS & CAPM
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (2)
Overview
1. Introduction:
Simple CAPM with quadratic utility functions
(from beta-state price equation)
2. Traditional Derivation of CAPM
– Demand: Portfolio Theory for given
– Aggregation: Fund Separation Theorem prices/returns
– Equilibrium: CAPM
3. Modern Derivation of CAPM
– Projections
– Pricing Kernel and Expectation Kernel
4. Testing CAPM
5. Practical Issues – Black-Litterman
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (3)
• ⇒ 𝜕1 𝑢 = −2 𝑥1,1 − 𝛼 , … , −2 𝑥𝑆,1 − 𝛼
• Excess return
ℎ 𝑓 ℎ
cov 𝑚, 𝑅 𝑅 cov 𝜕1 𝑢, 𝑅
𝐸 𝑅ℎ − 𝑅𝑓 = − =−
𝐸𝑚 𝐸 𝜕0 𝑢
𝑓
𝑅 cov −2 𝑥1 − 𝛼 , 𝑅 ℎ 2cov 𝑥 , 𝑅 ℎ
1
=− = 𝑅𝑓
𝐸 𝜕0 𝑢 𝐸 𝜕0 𝑢
• Also holds for market portfolio
𝐸 𝑅ℎ − 𝑅 𝑓 cov 𝑥1 , 𝑅ℎ
=
𝐸𝑅 𝑚𝑘𝑡 −𝑅 𝑓 cov 𝑥1 , 𝑅𝑚𝑘𝑡
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (5)
𝐸 𝑅ℎ − 𝑅 𝑓 cov 𝑥1 , 𝑅ℎ
=
𝐸 𝑅𝑚𝑘𝑡 − 𝑅 𝑓 cov 𝑥1 , 𝑅𝑚𝑘𝑡
Overview
1. Introduction:
Simple CAPM with quadratic utility functions
2. Traditional Derivation of CAPM
– Demand: Portfolio Theory for given
– Aggregation: Fund Separation Theorem prices/returns
– Equilibrium: CAPM
3. Modern Derivation of CAPM
– Projections
– Pricing Kernel and Expectation Kernel
4. Testing CAPM
5. Practical Issues – Black-Litterman
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (7)
– 𝜎ℎ2 ≔ var 𝑟ℎ = 𝒘′ 𝑉𝒘
𝜎12 𝜎12 𝑤1
= 𝑤1 𝑤2
𝜎21 𝜎22 𝑤2
= 𝑤12 𝜎12 + 𝑤22 𝜎22 + 2𝑤1 𝑤2 𝜎12 ≥ 0
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (9)
𝜇2
𝜇ℎ
𝜇1
𝜇2 − 𝜇1
𝜇ℎ = 𝜇1 + −𝜎ℎ − 𝜎1
𝜎2 − 𝜎1
±𝜎𝑝 +−𝜎2
For 𝜌12 = −1 ⇒ 𝑤1 =
𝜎1 +𝜎2
𝜎ℎ = 𝑤1 𝜎1 − 1 − 𝑤1 𝜎2
𝜎2 𝜎1 𝜇2 − 𝜇1
𝜇ℎ = 𝑤1 𝜇1 + 1 − 𝑤1 𝜇2 = 𝜇 + 𝜇 ± 𝜎
𝜎1 + 𝜎2 1 𝜎1 + 𝜎2 2 𝜎1 + 𝜎2 𝑝
𝜇2 𝜇 −𝜇
slope: 𝜎2 +𝜎1
1 2
𝜎2 𝜎
intercept: 𝜇 + 1 𝜇
𝜎1 +𝜎2 1 𝜎1 +𝜎2 2 𝜇 −𝜇
slope: − 𝜎2+𝜎1
𝜇1 1 2
𝜎1 𝜎2
E[r2]
E[r1]
s1 s2
For 𝜎1 = 0
𝜇2
𝜇ℎ
𝜇1
𝜎1 𝜎ℎ 𝜎2
𝜕ℒ
= 𝑉𝒘 − 𝜆𝝁 − 𝛾𝟏 = 0
𝜕𝑤
𝜕ℒ
= 𝜇ℎ − 𝒘′ 𝝁 = 0
𝜕𝜆
𝜕ℒ
= 1 − 𝒘′ 𝟏 = 0
𝜕𝛾
• Solving for 𝜆, 𝛾
𝐶𝜇ℎ − 𝐴 𝐵 − 𝐴𝜇ℎ
𝜆= , 𝛾=
𝐷 𝐷
𝐷 = 𝐵𝐶 − 𝐴2
skip
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (17)
• Hence, 𝒘ℎ = 𝜆𝑉 −1 𝝁 + 𝛾𝑉 −1 𝟏 becomes
𝐶𝜇ℎ − 𝐴 −1 𝐵 − 𝐴𝜇ℎ −1
𝒘ℎ = 𝑉 𝝁+ 𝑉 𝟏
𝐷 𝐷
1 1
= 𝐵 𝑉 𝟏 − 𝐴 𝑉 𝝁 + 𝐶 𝑉 −1 𝝁 − 𝐴 𝑉 −1 𝟏 𝜇ℎ
−1 −1
𝐷 𝐷
𝓰 𝓱
skip
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (18)
2
ℎ 𝐶 ℎ 𝐴 1
𝜎2 𝜇 = 𝜇 − +
𝐷 𝐶 𝐶
skip
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (20)
1
ii. the variance of the minimum variance portfolio is given by ;
𝐶
𝐶 𝐴 2 1
iii. Equation 𝜎2 𝜇ℎ = 𝜇ℎ − + is a
𝐷 𝐶 𝐶
1 𝐴
– parabola with vertex ,
𝐶 𝐶
in the expected return/variance space
– hyperbola in the expected return/standard deviation space.
skip
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (21)
𝐴 𝐷 2 1
𝐸 𝑟ℎ = ± 𝜎 −
𝐶 𝐶 𝐶
– Solution
𝑉 −1 𝝁−𝑟 𝑓 𝟏 𝜇 ℎ −𝑟 𝑓
• 𝒘ℎ = , where 𝐻 = 𝐵 − 2𝐴𝑟 𝑓 + 𝐶(𝑟 𝑓 )2
𝐻2
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (26)
2
𝐸 𝑟ℎ − 𝑟𝑓
var 𝑟ℎ =
𝐻2
𝐸 𝑟ℎ = 𝑟𝑓 + 𝐻𝜎ℎ
Overview
1. Introduction:
Simple CAPM with quadratic utility functions
2. Traditional Derivation of CAPM
– Demand: Portfolio Theory for given
– Aggregation: Fund Separation Theorem prices/returns
– Equilibrium: CAPM
3. Modern Derivation of CAPM
– Projections
– Pricing Kernel and Expectation Kernel
4. Testing CAPM
5. Practical Issues – Black-Litterman
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (29)
Price of Risk =
= highest
Sharpe ratio
Mean-Variance Preferences
𝜕𝑈 𝜕𝑈
• 𝑈 𝜇ℎ , 𝜎ℎ with > 0, 2 <0
𝜕𝜇ℎ 𝜕𝜎ℎ
𝜌
– Example: 𝐸 𝑊 − var 𝑊
2
Overview
1. Introduction:
Simple CAPM with quadratic utility functions
2. Traditional Derivation of CAPM
– Demand: Portfolio Theory for given
– Aggregation: Fund Separation Theorem prices/returns
– Equilibrium: CAPM
3. Modern Derivation of CAPM
– Projections
– Pricing Kernel and Expectation Kernel
4. Testing CAPM
5. Practical Issues – Black-Litterman
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (33)
CML
M
rM
rf
j
sM sp
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (36)
SML
E(ri)
E(rM)
rf
slope SML = (E(ri)-rf) /b i
b M= 1 bi b
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (37)
Overview
1. Introduction:
Simple CAPM with quadratic utility functions
2. Traditional Derivation of CAPM
– Demand: Portfolio Theory for given
prices/returns
– Aggregation: Fund Separation Theorem
– Equilibrium: CAPM
3. Modern Derivation of CAPM
– Projections
– Pricing Kernel and Expectation Kernel
4. Practical Issues
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (38)
Projections
• States 𝑠 = 1, … , 𝑆 with 𝜋𝑠 > 0
• Probability inner product
𝑥, 𝑦 𝜋 = 𝜋𝑠 𝑥𝑠 𝑦𝑠 = 𝜋𝑠 𝑥𝑠 𝜋𝑠 𝑦𝑠
𝑠 𝑠
• 𝜋-norm 𝑥 = 𝑥, 𝑥 𝜋 (measure of length)
i. 𝑥 > 0 ∀𝑥 ≠ 0 and 𝑥 = 0 if 𝑥 = 0
ii. 𝜆𝑥 = 𝜆 𝑥
iii. 𝑥 + 𝑦 ≤ 𝑥 + 𝑦 ∀𝑥; 𝑦 ∈ ℝ𝑆
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (39)
)
shrink
axes
y y
x x
…Projections…
• 𝒵 space of all linear combinations of vectors 𝑧1 , … , 𝑧𝑛
• Given a vector 𝑦 ∈ ℝ𝑆 solve
2
min𝑛 𝐸 𝑦 − 𝛼 𝑗 𝑧𝑗
𝛼∈ℝ
𝑗
𝑗 𝑗 𝑗
• FOC: 𝑠 𝜋𝑠 𝑦𝑠 − 𝑗 𝛼 𝑧𝑠 𝑧 = 0
– Solution 𝛼 ⇒ 𝑦 𝒵 = 𝑗 𝛼 𝑗 𝑧 𝑗 , 𝜖 ≔ 𝑦 − 𝑦 𝒵
…Projections
y
e
yZ
𝑥, 𝑦 = 𝐸 𝑥𝑦 = cov 𝑥, 𝑦 + 𝐸 𝑥 𝐸 𝑦
x
𝑥, 𝑥 = 𝐸 𝑥 2 = var 𝑥 + 𝐸 𝑥 2
𝑥 𝑥 = 𝐸 𝑥2
𝑥 =𝑥+𝑥
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (43)
– 𝜎𝑥 = 𝑥 𝜋
• cov 𝑥, 𝑦 = cov 𝑥, 𝑦 = 𝑥, 𝑦 𝜋
• Proof: 𝑥, 𝑦 𝜋 = 𝑥, 𝑦 𝜋 + 𝑥, 𝑦 𝜋
– 𝑦, 𝑥 𝜋 = 𝑦, 𝑥 𝜋 = 0, 𝑥, 𝑦 𝜋 = 𝐸 𝑦 𝐸 𝑥 + cov[𝑥, 𝑦]
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (44)
Overview
1. Introduction:
Simple CAPM with quadratic utility functions
2. Traditional Derivation of CAPM
– Demand: Portfolio Theory for given
– Aggregation: Fund Separation Theorem prices/returns
– Equilibrium: CAPM
3. Modern Derivation of CAPM
– Projections
– Pricing Kernel and Expectation Kernel
4. Testing CAPM
5. Practical Issues – Black-Litterman
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (45)
Pricing Kernel ∗
𝑚 …
• 𝑋 space of feasible payoffs.
• If no arbitrage and 𝜋 ≫ 0 there exists
SDF 𝑚 ∈ ℝ𝑆 , 𝑚 ≫ 0, such that 𝑞 𝑧 = 𝐸 𝑚𝑧 .
• 𝑚 ∈ ℝ𝑆 – SDF need not be in asset span.
• A pricing kernel is a 𝑚∗ ∈ 𝑋 such that for
each 𝑧 ∈ 𝑋 , 𝑞 𝑧 = 𝐸 𝑚∗ 𝑧
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (46)
Expectations Kernel 𝑘 ∗
– 𝑘 ∗ is the “projection” of 𝐼 on 𝑋
– 𝑘 ∗ = 𝐼 if bond can be replicated (e.g. if markets are complete)
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (49)
Frontier Returns…
• Frontier returns are the returns of frontier payoffs with non-zero
prices.
[Note: R indicates Gross return]
𝑘∗ 𝑘∗
𝑅𝑘 ∗ = =
𝑞∗ 𝑘 ∗ 𝐸𝑚 ∗
𝑚 𝑚∗
𝑅𝑚 ∗ = =
𝑞 𝑚∗ 𝐸 𝑚∗ 𝑚∗
• If 𝑧 = 𝛼𝑚∗ + 𝛽𝑘 ∗ then ∗
𝛼𝑞 𝑚 𝛽𝑞 𝑘 ∗
𝑅𝑧 = ∗ ∗ 𝑅𝑚 ∗ + ∗ ∗ 𝑅𝑘 ∗
𝛼𝑞 𝑚 + 𝛽𝑞 𝑘 𝛼𝑞 𝑚 + 𝛽𝑞 𝑘
𝜆 1−𝜆
𝑋 = RS = R3
𝑚∗
0 expected return
𝑚∗
0 expected return
𝑚∗
…Frontier Returns
(if agent is risk-neutral)
• If 𝑘 ∗ = 𝛼𝑚∗ , frontier returns ≡ 𝑅𝑘 ∗
• If𝑘 ∗ ≠ 𝛼𝑚∗ , frontier returns can be written as:
𝑅𝜆 = 𝑅𝑘 ∗ + 𝜆 𝑅𝑚∗ − 𝑅𝑘 ∗
• Expectations and variance are
𝐸 𝑅𝜆 = 𝐸 𝑅𝑘 ∗ + 𝜆 𝐸 𝑅𝑚∗ − 𝐸 𝑅𝑘∗
var 𝑅𝜆 =
= var 𝑅𝑘 ∗ + 2𝜆cov 𝑅𝑘 ∗ , 𝑅𝑚∗ − 𝑅𝑘 ∗ + 𝜆2 var 𝑅𝑚∗ − 𝑅𝑘 ∗
• If risk-free asset exists, these simplify to:
𝐸 𝑅𝑚∗ − 𝑅𝑓
𝐸 𝑅𝜆 = 𝑅𝑓 + 𝜆 𝐸 𝑅𝑚∗ − 𝑅𝑓 = 𝑅𝑓 ± 𝜎 𝑅𝜆
𝜎 𝑅𝑚∗
var 𝑅𝜆 = 𝜆2 var[𝑅𝑚∗ ] , 𝜎 𝑅𝜆 = 𝜆 𝜎 𝑅𝑚∗
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (55)
Illustration of MVP
𝑋 = ℝ2 and 𝑆 = 3
Expected return
of MVP
Minimum standard
deviation
(1,1,1)
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (57)
Illustration of ZC Portfolio…
𝑋 = ℝ2 and 𝑆 = 3
arbitrary portfolio p
(1,1,1)
Recall:
cov 𝑥, 𝑦 = 𝑥, 𝑦 𝜋
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (60)
…Illustration of ZC Portfolio
Green lines do not
necessarily cross.
arbitrary portfolio p
(1,1,1)
ZC of p
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (61)
Beta Pricing…
• Frontier Returns (are on linear subspace). Hence
𝑅𝛽 = 𝑅𝜇 + 𝛽 𝑅𝜆 − 𝑅𝜇
• Consider any asset with payoff 𝑥𝑗
– It can be decomposed in 𝑥𝑗 = 𝑥𝑗𝜀 + 𝜀𝑖
– 𝑞 𝑥𝑗 = 𝑞 𝑥𝑗𝜀 and 𝐸 𝑥𝑗 = 𝐸 𝑥𝑗𝜀 , since 𝜀 ⊥ ℰ
𝜀𝑗
– Return of 𝑥𝑗 is 𝑅𝑗 = 𝑅𝑗𝜀 +
𝑞 𝑥𝑗
– Using above and assuming 𝜆 ≠ 𝜆0 and 𝜇 is
ZC-portfolio of 𝜆,
𝜀𝑗
𝑅𝑗 = 𝑅𝜇 + 𝛽𝑗 𝑅𝜆 − 𝑅𝜇 +
𝑞 𝑥𝑗
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (62)
…Beta Pricing
• Taking expectations and deriving covariance
• 𝐸 𝑅𝑗 = 𝐸 𝑅𝜇 + 𝛽𝑗 𝐸 𝑅𝜆 − 𝐸 𝑅𝜇
cov 𝑅𝜆 ,𝑅𝑗
• cov 𝑅𝜆 , 𝑅𝑗 = 𝛽𝑗 var 𝑅𝜆 ⇒ 𝛽𝑗 =
var 𝑅𝜆
𝜀𝑗
– Since 𝑅𝜆 ⊥
𝑞 𝑥𝑗
• If risk-free asset can be replicated, beta-pricing
equation simplifies to
𝐸 𝑅𝑗 = 𝑅𝑓 + 𝛽𝑗 𝐸 𝑅𝜆 − 𝑅𝑓
• N.B. first equation always hold if there are only two assets.
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (65)
Overview
1. Introduction:
Simple CAPM with quadratic utility functions
2. Traditional Derivation of CAPM
– Demand: Portfolio Theory for given
– Aggregation: Fund Separation Theorem prices/returns
– Equilibrium: CAPM
3. Modern Derivation of CAPM
– Projections
– Pricing Kernel and Expectation Kernel
4. Testing CAPM
5. Practical Issues – Black-Litterman
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (66)
Practical Issues
• Testing of CAPM
• Jumping weights
– Domestic investments
– International investment
• Black-Litterman solution
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (67)
Empirical Evidence
• Excess returns on high-beta stocks are low
• Excess returns are high for small stocks
– Effect has been weak since early 1980s
• Value stocks have high returns despite low
betas
• Momentum stocks have high returns and low
betas
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (69)
Practical Issues
• Estimation
– How do we estimate all the parameters we need for
portfolio optimization?
Overview
1. Introduction:
Simple CAPM with quadratic utility functions
2. Traditional Derivation of CAPM
– Demand: Portfolio Theory for given
– Aggregation: Fund Separation Theorem prices/returns
– Equilibrium: CAPM
3. Modern Derivation of CAPM
– Projections
– Pricing Kernel and Expectation Kernel
4. Testing CAPM
5. Practical Issues – Black Litterman
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (72)
• Estimating means
– For any partition of [0,T] with N points (∆t=T/N):
1 1 𝑁 𝑝𝑇 −𝑝0
𝐸𝑟 = ⋅ ⋅ 𝑖=1 𝑟𝑖⋅Δ𝑡 = (in log prices)
Δ𝑡 𝑁 𝑇
– Knowing the first and last price is sufficient
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (73)
Estimating Means
• Let 𝑋𝑘 denote the logarithmic return on the market, with 𝑘 =
1, … , 𝑛 over a period of length ℎ
– The dynamics to be estimated are:
𝑋𝑘 = 𝜇 ⋅ Δ + 𝜎 ⋅ Δ ⋅ 𝜖𝑘
where the 𝜖𝑘 are i.i.d. standard normal random variables.
– The standard estimator for the expected logarithmic mean rate of
return is:
1 𝑛 where
𝜇= ⋅ 𝑋𝑘 ℎ is length of observation
ℎ 1
𝑛 number of observations
– The mean and variance of this estimator Δ = 𝑛/ℎ
1 𝑛 1
𝐸 𝜇 = ⋅𝐸 𝑋𝑘 = ⋅𝑛⋅𝜇⋅Δ=𝜇
ℎ 1 ℎ
1 𝑛 1 2
𝜎2
𝑉𝑎𝑟 𝜇 = 2 ⋅ 𝑉𝑎𝑟 𝑋𝑘 = 2⋅𝑛⋅𝜎 ⋅Δ=
ℎ 1 ℎ ℎ
– The accuracy of the estimator depends only upon the total length of the observation period (h), and not upon the number of
observations (n).
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (74)
Estimating Variances
• Consider the following estimator:
1 𝑛 2
𝜎2 = ⋅ 𝑖=1 𝑋𝑘
ℎ
Estimating variances:
Theory vs. Practice
• For any partition of [0, 𝑇] with 𝑁 points (Δ𝑡 = 𝑇/𝑁):
𝑁
1 2
𝑉𝑎𝑟 𝑟 = ⋅ 𝑟𝑖⋅Δ𝑡 − 𝐸 𝑟 → 𝜎 2 𝑎𝑠 𝑁 → ∞
𝑁
𝑖=1
Estimating covariances:
Theory vs. Practice
• In theory, the estimation of covariances shares
the features of variance estimation.
• In practice:
– Difficult to obtain synchronously observed time-series -> may require
interpolation, which affects the covariance estimates.
– The number of covariances to be estimated grows very quickly, such
that the resulting covariance matrices are unstable (check condition
numbers!).
– Shrinkage estimators (Ledoit and Wolf, 2003, “Honey, I Shrunk the
Covariance Matrix”)
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (77)
Britton-Jones (1999)
1977-1996 1977-1986 1987-1996
weights t-stats weights t-stats weights t-stats
Australia 12.8 0.54 6.8 0.20 21.6 0.66
Austria 3.0 0.12 -9.7 -0.22 22.5 0.74
Belgium 29.0 0.83 7.1 0.15 66 1.21
Canada -45.2 -1.16 -32.7 -0.64 -68.9 -1.10
Denmark 14.2 0.47 -29.6 -0.65 68.8 1.78
France 1.2 0.04 -0.7 -0.02 -22.8 -0.48
Germany -18.2 -0.51 9.4 0.19 -58.6 -1.13
Italy 5.9 0.29 22.2 0.79 -15.3 -0.52
Japan 5.6 0.24 57.7 1.43 -24.5 -0.87
UK 32.5 1.01 42.5 0.99 3.5 0.07
US 59.3 1.26 27.0 0.41 107.9 1.53
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (79)
Black-Litterman Appraoch
• Since portfolio weights are very unstable, we need to discipline our
estimates somehow
– Our current approach focuses only on historical data
• Priors
– Unusually high (or low) past return may not (on average) earn the same
high (or low) return going forward
– Highly correlated sectors should have similar expected returns
– A “good deal” in the past (i.e. a good realized return relative to risk)
should not persist if everyone is applying mean-variance optimization.
• The investor is assumed to start with the following Bayesian prior (with
imprecision):
𝜇 = Π + 𝜖 𝑒𝑞 where 𝜖 𝑒𝑞 ∼ 𝑁 0, 𝜏 ⋅ Σ
– The precision of the equilibrium return estimates is assumed to be
proportional to the variance of the returns.
– 𝜏 is a scaling parameter
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (81)
• Posterior distribution:
– If 𝑋1 , 𝑋2 are normally distributed as:
𝑋1 𝜇1 Σ Σ12
∼ 𝑁 𝜇 , 11
𝑋2 2 Σ21 Σ22
– Then, the conditional distribution is given by
−1 −1
𝑋1 |𝑋2 = 𝑥 ∼ 𝑁 𝜇1 + Σ12 Σ22 𝑥 − 𝜇2 , Σ11 − Σ12 Σ22 Σ21
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (83)
−1
var 𝑅 𝑄 = 𝜏⋅Σ + 𝑃′ ⋅ Ω−1 ⋅ 𝑃 −1
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (84)
−1
−1
1
var 𝑅 𝑄 = 𝜏⋅Σ +
2Ω
FIN501 Asset Pricing
Lecture 06 Mean-Variance & CAPM (85)
Advantages of Black-Litterman
• Returns are adjusted only partially toward the
investor’s views using Bayesian updating
– Recognizes that views may be due to estimation error
– Only highly precise/confident views are weighted heavily.
• Returns are modified in way that is consistent with
economic priors
– Highly correlated sectors have returns modified in the
same direction.
• Returns can be modified to reflect absolute or relative
views.
• Resulting weight are reasonable and do not load up on
estimation error.